Affiliate Commission Tiers: The Best Way to Reward Your Affiliates

Good affiliates are the most portable people in your program.

They talk to other brands, they get offers, and in a typical store that one person accounts for more than half of your referred revenue.

If they go, most of the channel goes with them.

Meanwhile the affiliate who could become your next good one has nothing to aim at, because doing well and doing nothing pay the same.

Commission tiers fix both halves of that.

You pay more only to the people who have already proved they sell, and you publish the rungs so everyone else can see what climbing is worth.

Here is the shape of a program that is actually producing:

• One affiliate is producing most of your referred sales.
• A few others have sold once or twice.
• And a long list of people signed up and never did anything at all.

So how do you reward the ones who are actually producing, without handing more money to everyone who is not?

There are five ways merchants usually try.

Four of them fail, and they fail for the same underlying reason.

The fifth is commission tiers.

Worth knowing before we start: in our data across 1,178 Shopify affiliate programs, 59% of affiliates are already paid through a tiered structure.

Tiers are not an advanced tactic.

They are the norm, and most of them do nothing, which is a problem of design rather than of the idea.

The five ways to reward affiliates

Raise everyone's rate

The instinct, and the weakest option.

Across 471 programs we found no meaningful correlation between the commission rate a store sets and how that program performs.

It also misses by construction: roughly one affiliate in six ever makes a sale, so a blanket rise pays the other five nothing extra for nothing, since zero at 15% is the same as zero at 10%.

Pay a one-off bonus

Genuinely useful, and the right tool for a launch or a seasonal push. But it rewards a moment rather than a pattern, it has to be re-decided every time, and it gives an affiliate no reason to plan around you next quarter.

Send gifts or free product

Good for relationships and bad as compensation.

It does not scale past a handful of people, it costs you real COGS with no attribution attached, and an affiliate who is treating this as income cannot pay rent with a hoodie.

Give a custom rate to one person

This works, and for a single standout it is the right answer.

Its limit is that it is invisible. Nobody else in your program knows it exists, so it rewards one affiliate without giving anyone else a reason to aim higher.

Build a tier ladder

An affiliate earns a higher rate as their own sales pass thresholds you set in advance.

Besides the data, I believe merchants already believe that a tiered system works, because before we added our own tiered feature on Affilitrak, it was one of the most requested features at the time.

Why tiers beat the other four

Look at what the first four have in common.

Each one either pays people who are not producing, or rewards a single moment, or rewards one person privately.

A ladder does none of those things.

• It is targeted: Only affiliates who have already proved they sell reach a higher rate, and they reach it out of revenue that already exists. You never pay more in advance.
• It is self-serve: Once the rungs are published, the program rewards performance without you making a decision each time. No negotiation, no case-by-case judgement, no awkward conversation with the affiliate you did not upgrade.
• It is visible: This is the part the custom rate cannot do. Every affiliate can see what the next rung pays and what it takes to get there, which turns a vague sense that more is better into a specific target.
• It compounds trust: A published ladder tells affiliates the rules are the same for everyone. Programs that pay quietly different rates for unstated reasons lose good people to the ones that do not.

And there is a specific person it exists to keep.

One affiliate carries most of the program, however many you recruit

Selling affiliates in the store	Top affiliate's share of orders
2 or more	54.5%
3 or more	44.6%
5 or more	35.1%
10 or more	26.4%

In stores with at least two selling affiliates, the top one takes a median 54.5% of referred orders.

That person is not one of your affiliates. They are your program.

A ladder is the cheapest insurance available against them leaving, and it costs you nothing until they have already earned it.

One thing to clear up first

The word "tiers" covers two unrelated mechanisms, and they get sold under the same label.

• Performance tiers: an affiliate earns a higher rate as their own sales grow. 10%, then 12.5%, then 15%. This is the ladder, and it is what this article is about.
• Multi-level commissions: an affiliate earns a cut of what affiliates they recruited produce. That is a recruitment structure, not a reward for performance, and it carries different considerations depending on where you operate.

If you are trying to reward your best seller, you want the first one.

How to set the tiers

Set tier one where an affiliate already stands.

Do not pick 50 orders because it is a round number.

Open your dashboard, sort by referred sales, and put the first tier just above where your second or third best performer already sits.

Somebody crosses it in the first month, and the ladder becomes visibly real to everyone else.

A threshold nobody has ever reached teaches your affiliates that the ladder is decoration.

Measure revenue, not order count

Orders reward whoever sends the most cheap items.

Referred revenue rewards whoever sends the most value, which is what you actually care about.

If everything you sell is one price, the order count is fine.

Set a maximum of three tiers

Three tiers is the most you should run, and that's mostly based on the tiered programs I've helped merchants set.

It should consist of:

• A base rate
• A proven-seller rate
• A top-performer rate

Every extra tier makes the program harder to explain and harder to trust.

Make each step big enough to notice

A jump from 10% to 11% motivates nobody and still costs you margin, which is the worst of both.

Something like 10% to 12.5% to 15% reads as a real ladder.

Decide the measurement window and say it out loud

Lifetime totals make the ladder a ratchet: once earned, never lost, and a new affiliate can never catch someone who joined two years ago.

A rolling 30 or 90 day window keeps tiers current but means people can fall.

Both are defensible. Silence is not, because an affiliate who drops a rung without warning assumes you cut their rate.

Cap the top rung with your margin

Work out what your margin can carry, then set your highest tier at or below it.

If 15% is your ceiling, your base has to start low enough that there is somewhere to climb to.

Who a ladder actually moves

Be honest about this, because it changes how you judge whether it is working.

• It does not move the 84% of affiliates who never sell. No commission structure does. That is a recruiting problem.
• It moves the middle. The affiliate who has made four sales and could make ten. This is where a ladder earns its keep.
• It retains the top. Not by pushing them to sell more, since they are likely already at capacity, but by making leaving expensive.

So measure a ladder by how many affiliates climb a rung, not by total revenue. If nobody has moved up in three months, your tiers are in the wrong place.

The five ways ladders fail

• Tiers nobody reaches: The most common failure, and it always traces back to round numbers chosen before looking at the data.
• Lifetime thresholds that lock newcomers out: If your best affiliate cleared the top rung two years ago and nobody can catch them, the ladder stopped working for everyone else.
• Increments too small to register: One percentage point is a rounding error to an affiliate and a real cost to you.
• Tiers nobody was told about: An affiliate who does not know the next rung exists cannot aim at it. It belongs in the welcome email, not buried in terms.
• Forgetting that commissions reverse: If a refund drops someone below a threshold, decide in advance whether they keep the rung for that period. Deciding afterwards looks like moving the goalposts.

When a ladder is the wrong tool

Two cases where something simpler fits:

• You have one selling affiliate and nobody else: A ladder needs a middle group to pull upward. Give that person a custom rate directly and spend the energy recruiting a second seller.
• You want a short-term push: For a launch or a seasonal window, a one-off bonus is cleaner than permanently restructuring your rates.

Both tiered structures and per-affiliate rates are standard in Affilitrak, set per program or per individual with no upgrade required.

That matters more than it sounds, because on several affiliate apps tiered commissions sit behind the second or third pricing tier, so your ladder costs you a subscription increase before it pays anyone anything.

The Programs guide walks through how to set a ladder in Affilitrak if you want the steps.

Conclusion

Every other way of rewarding affiliates pays the wrong people, rewards a single moment, or rewards one person in private.

A ladder is the only one that is targeted, self-serve, visible to everyone, and funded by revenue that already exists.

But it only works if the tiers are in reachable places.

Build yours from your own numbers.

First tier just above your second-best performer, revenue rather than order count, three steps at most, each step big enough to notice, and every affiliate told exactly where the tiers are and over what period.

Then judge it by climbers, not by revenue.

And keep it in proportion.

A ladder grows and keeps the affiliates who sell.

It will not create them.

If nobody is near your second rung yet, the problem is recruiting, and our guide to setting up an affiliate program on Shopify is the better place to start.

You can install Affilitrak free and set tiered or per-affiliate rates without upgrading anything.